Here is the page of my personal build Mozilla Firefox 2.0. In this version, it adopts a method to bypass the Websense content filter. Just read the details in the page. We have Windows 32bit version binary and Mac OS X PowerPC G4 (7450) version binary of Firefox. For somebody will still encounter browsing difficulities behind Websense network, just adjust some perferences in Firefox (about:config).

1. set network.http.keep-alive to false
2. set network.http.max-persistent-connections-per-server to 0

Please refer to this page for details.

It is very easy to modify a HTTP proxy server to adopt Websense bypass mechanise. Take Muffin for example, it is an open-source HTTP/HTTPS proxy server written in Java programming language. In src/org/doit/muffin/Request.java, write() method, the original code,

super.write(out);
if (data != null)
{
out.write(data);
out.flush();
}

Just have some modification,

ByteArray ba = super.toByteArray();
byte[] array = ba.getBytes();
out.write( array, 0, 1 ); out.flush();
out.write( array , 1, 1 ); out.flush();
out.write( array, 2, ba.length()-2); out.flush();
//super.write(out);
if (data != null)
{
out.write(data);
out.flush();
}

The new code will force HTTP request to become multiple segmented packets. For example, GET http://tw.yahoo.com HTTP/1.1 will become "G" in one packet, "E" in another packet, and T http://tw..... to the end of request will be in other packets. By this way, Websense will not able to get the FULL HTTP request packet. Therefore, Websense will not able to block you out.

Tags: MacOS X, websense

In Tech-Faq , there is an article talking about how to by pass Websense. Instead of using proxy or tunneling methods, there is still a way to bypass the firewall filter. In the article, there are some important sentences, listed in following,

In transparent mode, Websense counts on the fact that the firewall will forward the whole request in a single time. If the whole request is not transferred at one time, Websense will allow the packet to pass, as the packet doesn't look like a HTTP request.

In the fact, if you browse a prohibited website, the Websense server will feedback an HTTP redirect header to your browser. Then your browser will go to the page like this. The techniques of above sentences could clearly explain why I can bypass the Websense via Telnet. Instead of using any browsers, I just telneted to a prohibited website host with port 80. Then I keyed in the HTTP protocol strings, for example, GET /index.html HTTP/1.1. Then The raw HTML content was dumped to the terminal, not Websense HTTP redirect or blocked page.
Maybe I can modify the PERL HTTP proxy program I wrote before to adopt this mechanism to bypass the Websense. :P Because of the ridiculous policies and reasons of preventing PCs from attacks of viruses, Trojans, worms, and hackers, they blocked out some sites like Flickr, wretch, pixnet, xuite, google spreadsheets, and so on. It is very ridiculous and amazing.

Tags: MacOS X, websense